The Principle of Least Action from the Stationary Phase Approximation

Hello! Welcome back to our journey through the path integral formulation of quantum mechanics.

In our last lesson, we successfully calculated the exact propagators for the free particle and the quantum harmonic oscillator. We saw that for these systems with quadratic Lagrangians, the propagator takes the form $K\sim e^{i{S}_{cl}/\hslash }$, where the phase is determined entirely by the classical action.

However, most physical systems, especially those with interesting interactions in condensed matter, have Lagrangians that are not quadratic. This means their path integrals cannot be solved exactly. Today, we'll learn the primary tool for tackling these more complex cases: the stationary phase approximation. Our goal is to apply this method to the path integral to understand how classical mechanics, in the form of the Principle of Least Action, emerges as a limit of the more fundamental quantum theory. This is one of the most profound and beautiful connections in all of physics.

The Interference of Paths

The path integral tells us to sum up contributions from all possible paths a particle can take between two points. Each path $x(t)$ contributes a complex number, or "phasor," $e^{iS[x(t)]/\hslash }$, where $S[x(t)]$ is the action for that path.

The key to understanding the emergence of classical mechanics lies in the denominator of the phase: Planck's constant, $\hslash$. For a macroscopic object like a baseball, the typical action $S$ is enormous compared to $\hslash \approx 10^{-34}$ J·s. This means the phase, $S/\hslash$, is an astronomically large number.

What happens when we compare two paths that are very close to each other? Even a tiny change in the path leads to a small change in the action, $\delta S$. But the phase changes by $\delta S/\hslash$, which is a huge number. Consequently, the phasors for these two nearby paths will point in completely different directions in the complex plane. When we sum up the contributions from a whole collection of such "non-classical" paths, their phasors point in all directions and destructively interfere, averaging to zero.

So, where does a non-zero answer come from? It comes from a special region of paths where this cancellation doesn't happen. The following video provides a superb visual and intuitive explanation of this phenomenon.

The Path of Stationary Action

As the video explained, the destructive interference fails for one special path (and its immediate neighbors) where the action is stationary. This means that for this particular path, a small variation produces no first-order change in the action.

For paths in this neighborhood, the action is nearly constant. Therefore, their corresponding phasors, $e^{iS/\hslash }$, all point in almost the same direction. They interfere constructively, producing a dominant contribution to the path integral.

In the classical limit ($\hslash \to 0$), this constructive interference is so sharply peaked that the entire path integral is determined by the contribution from this single path of stationary action.

Recovering the Principle of Least Action

You've certainly encountered the principle of stationary action (often called the Principle of Least Action) in your classical mechanics courses. It states that the trajectory a classical particle follows between two points is the one that makes the action stationary. What is truly remarkable is that we can prove this principle is mathematically identical to Newton's laws of motion.

The following video lecture gives a concise proof of this equivalence. Your familiarity with the calculus of variations will be very helpful here.

By working through the variation $\delta S=\delta \int L(x,\dot{x})dt=0$ and using integration by parts, one arrives at the Euler-Lagrange equation:

For a standard Lagrangian $L=\frac{1}{2}m{\dot{x}}^2-V(x)$, this becomes $-\frac{\partial V}{\partial x}-m\ddot{x}=0$, which is precisely Newton's second law, $F=ma$.

The path integral formulation thus provides a deep physical explanation for Hamilton's principle: the principle of least action works in classical mechanics because of the constructive and destructive interference of quantum probability amplitudes.

The Stationary Phase Approximation in Practice

We can formalize this idea into a powerful mathematical tool. The stationary phase approximation provides an explicit formula for an integral with a rapidly oscillating phase. When applied to the path integral, it gives us the semiclassical approximation to the propagator.

The following reading explains the method in detail, first for an ordinary integral and then for the full path integral.

The key result of this approximation is that the propagator can be written as a sum over all possible classical paths connecting the initial and final points:

Each term has two parts:

1. A phase factor $e^{i{S}_{cl}/\hslash }$, determined by the action along that classical path.
2. An amplitude $A_{cl}$, which arises from a Gaussian integral over the small quantum fluctuations around that classical path. This amplitude depends on the second variation of the action, ${\delta }^{2}S$.

This is a beautiful generalization of what we found last lesson. For quadratic systems, there is only one classical path, and this semiclassical approximation becomes exact. For general systems, it provides the leading contribution in the limit where $S\gg \hslash$.

Conclusion

Today, we bridged the gap between the quantum and classical worlds. By analyzing the interference of paths in the Feynman path integral, we saw how classical reality emerges from quantum rules.

Key Takeaways:

• The path integral is a sum over phasors $e^{iS/\hslash }$ for all possible paths.
• In the classical limit ($S\gg \hslash$), the phase oscillates extremely rapidly. Paths that are not "stationary" destructively interfere and cancel each other out.
• The path of stationary action ($\delta S=0$) and its immediate neighbors interfere constructively, dominating the integral.
• The condition $\delta S=0$ is precisely Hamilton's Principle, which is equivalent to the classical equations of motion (e.g., Newton's second law).
• The path integral formulation provides a deep physical reason for the Principle of Least Action: it is a direct consequence of the wave-like nature of quantum mechanics.

Preview of the Next Lesson:
So far, we have worked exclusively in real time, $t$. A fascinating and powerful technique involves performing an analytical continuation to imaginary time by making the substitution $t\to -i\tau$. This is known as a Wick rotation. We will see that this transforms the oscillating quantum phase factor $e^{iS/\hslash }$ into a real, decaying exponential $e^{-S_E/\hslash }$, where $S_E$ is the "Euclidean action." This seemingly abstract trick creates a profound connection between quantum mechanics and statistical mechanics, relating the path integral to the partition function. This connection is fundamental in many areas of condensed matter and quantum field theory.